Andrew Simms, SGR, charts a potentially lethal alignment of wealth, power and political interests in the fossil fuel and information industries, with AI entrenching and accelerating existing bias and inequalities.
Article from Responsible Science journal no. 8, April 2026. Published online 20 April 2026.
Billionaire media owners fly in and out of Silicon Valley on private jets, while they slyly insinuate Artificial Intelligence (AI) into our lives. We know it matters, and that there are consequences for democracy, mental health and the biosphere. Yet it can seem so remote, the work of the polluter and cyber elites is hard to pin down.
Then you get home from the pub, where you’d been laughing with a friend imagining a smart, robot vacuum cleaner with legs that meant it could climb stairs, and tap into your social media. A chill goes down your back when you are quickly confronted with an advert for a stair-climbing robot vacuum cleaner…
Real or urban myth? Many have had the experience of feeling targeted by adverts for something they’re convinced they’ve only ever just mentioned in conversation. Are we being listened to?
Corporate capitalism’s eagerness to keep us profitably over-consuming means that yes, sometimes, we are, in fact, or have been listened to.
In January 2025 tech giant Apple agreed a $95 million lawsuit to settle allegations that it used its AI assistant to eavesdrop on the conversations of customers using its products, and that voice recordings it made were shared with advertisers.
Online giant Amazon was fined over $30 million over the misuse of recordings made by its ‘always listening’ AI assistant, Alexa. The US Federal Trade Commission’s Bureau of Consumer Protection said Amazon was guilty of "misleading parents, keeping children's recordings indefinitely, and flouting parents' deletion requests," and had "sacrificed privacy for profits".
Meanwhile, in early 2026 Google paid $68 million to settle a class action lawsuit that was brought over its devices both listening to and recording people’s private conversations.
Digital eavesdropping
This may seem bad enough, that the monopolistic firms of the modern, digital age are all being found guilty of breaking the law with a form of ‘surveillance capitalism’. But the apparently legal, and real reason that a stair-climbing robot vacuum cleaner turns up on your feed after coming up in conversation is potentially even more insidious and disturbing.
Where the other big beast, Meta, is concerned, here’s how it works with regard to their platforms Facebook and Instagram, according to US ad agency McNutt & Partners. It’s down to an incredibly sophisticated form of tracking that converges multiple data sources about not just you, but the people you hang out with - in effect, digital eavesdropping, or spying.
Assume you never texted about or searched online for the product, but only spoke about it, and it’s now staring up at you from your screen. What happened? We already know that Facebook tracks your online searches and which websites you visit. Many, however, will be unaware or have forgotten that with the right box ticked or unticked, even if you’re not using the Facebook app, it is tracking your location.
That means it knows who you’re mixing with, and if the person you were speaking to has looked up, or had any interaction regarding the product online, then that will make the connection. But that ad could show up even if you never discussed the product, but your friend did with one of their friends. Because Facebook knows you were together it will deduce that this third party connection is enough to push the ad to you. As the agency puts it, ‘Facebook’s algorithm compares your interests, demographics, places you’ve been, groups you’re a part of, hashtags you follow (the list goes on) to that of your friend,’ and where your unique digital identities converge sufficiently you’ll be targeted.
You are the product
But it can go further still, to the point where it seems you have only ‘thought’ about something and an ad appears. Your thoughts are a stream of consciousness with one thing leading to another, connected in ways you might not be aware of. Facebook’s algorithm combines with the vast amount of data that constitutes your digital ID in not unsimilar ways, able to see patterns and make connections to scraps of information you may have left behind before or after having the thought, and speculatively offering up the ad as consequence. When this happens, it is likely not a coincidence.
To what extremes AI will take this ‘surveillance capitalism’ is hard to predict, like the technology itself. But in the way that early advertising practitioners styled themselves, this is ‘consumption engineering’ in which you, in the form of your personal data, are being sold to advertisers, in ways that makes the imperative and pressure of consumerism increasingly impossible to escape.
Consequences can be seen in how humanity as a whole goes into Earth Overshoot ever earlier in the year, by 24th July in 2025, and that seven of nine identified planetary ecological boundaries have now been transgressed due to overconsumption and the linked generation of waste.
Before AI came along this trajectory was already established, the danger is that AI will accelerate these trends, putting overconsumption on steroids.
Research by the Hot or Cool Institute found that the global average lifestyle carbon footprint across 25 high, upper middle and lower middle income countries is more than six times the 1.5°C-aligned target, 7.1 tCO₂e per person per year compared to 1.1 tCO₂e, the target for the year 2035. However, globally there is a polluter elite. The United States, for example, has an average lifestyle carbon footprint of 18.1 tonnes of carbon dioxide equivalent (tCO2e), more than 10 times that in Nigeria (1.5 tCO₂e). Three areas - food, housing and transport – are the main drivers of lifestyle-related emissions, together making up 66% to 95% of the total lifestyle carbon footprint. But AI looks set to supercharge things.
Energy hungry
According to the International Energy Agency, “Electricity demand from data centres worldwide is set to more than double by 2030 to slightly more than the entire electricity consumption of Japan. AI will be the most significant driver with electricity demand from AI-optimised data centres projected to more than quadruple by 2030.” It points out that in the United States, “data centres are set to account for almost half of the growth in electricity demand between now and 2030,” and “Driven by AI use, the US is set to consume more electricity in 2030 for processing data than for manufacturing all energy-intensive goods combined, including aluminium, steel, cement and chemicals.”
Dr Simon Clark, the atmospheric physicist and science communicator, speaking at SGR’s annual Responsible Science Conference in 2025, said that the corporations pushing AI, “Are, in many ways, the climate crisis in microcosm, because they exemplify this idea of… growth at all costs.” He went on to say that, “These AI companies are operating in an under-regulated space, where the environmental costs of their products are treated as a negative externality.”
In this context, existing locked-in imbalances of power look set to increase inequalities, as those best placed to exploit and benefit from the technology do so.
The Centre for Global Development argues that this is so for three major reasons. Richer countries are better simply better equipped to exploit the tools of AI, and poorer countries have less capacity to handle the technology’s disruptions. AI too is a particular threat to traditional manufacturing-led development models with, for example, as many as 60 percent of jobs in Bangladesh’s garment sectors potentially being lost to automation by 2030.
Also speaking at the Responsible Science Conference was Emily Ghosh, Senior Scientist and Program Director of the Equitable Transitions program at the Stockholm Environment Institute US. Ghosh spoke of the, “new layer of inequality emerging with the rise of these cyber elites”, made up of tech giants and billionaires who dominate even the emerging green tech economy who are, “not just shaping how we’re using data, but how data, resources, and even our personal behaviours are monetised.” Noting that the tech industry's role in the low carbon transition is often presented as a solution, she added that, “While Google, Amazon and Tesla are making massive investments in renewable energy, electric vehicles and AI, they have disproportionate control over many aspects of our lives, or information, of resources, and labour.”
Driving inequality
AI could also be an engine of domestic inequality, negatively reinforcing existing biases, with examples ranging from AI algorithms rejecting up to 80% of mortgage applications from Black families in the US, to an AI system downgrading exam results during the COVID-19 pandemic for 39% of students, disproportionately hitting pupils from disadvantaged schools.
Richer people pollute more, due to their increased consumption of high carbon products and services - flights, big houses and large cars, possibly several. Some estimates show that if issues of ownership and control are not addressed, even with necessary investments in climate action, the global wealth share of the top one percent could rise from around 38.5% today to 46% in 2050, an asymmetric distribution of benefits that fuels inequality and is a driver of both division, poorer well being and further over consumption.
The Media Reform Coalition highlights the concentration of media ownership which also represents a centralisation of power and concentration of wealth. Nearly half, seven of the top 15 online platforms used to access news in the UK are owned by Meta, Alphabet (owners of Google) and X Corp (formerly Twitter). Google has captured an astonishing 93% of UK search engine use. Advertising revenue is the major driver for both, with Meta (Facebook and Instagram) and Google combined accounting for around 60% of all UK advertising spend.
Three companies alone - DMG Media, News UK and Reach - control 90% of the national newspaper market (which also represents an important online presence) - a sector that has seen a 20% rise in market concentration since 2024. And, in terms of streaming, the market is enormously centralised too, with Netflix, Amazon Prime and Disney+ making up 75% of all UK subscriptions.
When these media owners lend their support to politicians who oppose regulation of industry and redistribution of wealth, it represents a convergence of interests with profound, one-sided political and climate consequences.
Power imbalance
In 2024, Alphabet, Meta, Microsoft and Amazon all allegedly supported US election candidates who were climate deniers. Elon Musk spent a reported $277 million backing Trump and other Republican candidates for election.
Apple, Microsoft and Google’s chief executives all offered “congratulations” to Donald Trump, for his second swearing-in as president and gave millions of dollars to his inaugural committee.
The Facebook founder, Mark Zuckerberg, said he was,“Optimistic and celebrating,” as he prepared to jointly host a ball for Trump’s inauguration, in comments that came minutes after Trump had ordered the nation’s exit from the Paris Agreement.
Trump’s administration went on to announce a range of policy recommendations for AI that included weakening the Clean Water Act in connection to water-hungry data centres. The proposals were reported to closely echo recommendations suggested by Meta, and a lobbying group representing the interests of big data centres run by the likes of Google and Amazon.
Political bias seems increasingly designed into the technology itself, with polarising algorithms propelling the ‘addiction engines’ of social media platforms. Research by Finnish think-tank Sitra into material fed to young people on Instagram, X and TikTok revealed it carried a strong right wing bias. Profiles in the 18-24 age group were studied across three countries, France, Finland and Romania. Content served to them by the algorithm was found to be 58% right wing, 16% centrist and only 26% to the left. Earlier research on the platform rebranded from Twitter to X by Elon Musk, shows an increase of right wing bias in terms of material promoted shortly after Musk’s endorsement of Donald Trump in July 2024.
Internal documents released by whistleblowers from Meta and TikTok also reveal a dangerous alignment between corporations’ commercial objectives and the promotion of extreme, racist, sexist and violent material - all of which pander to right wing, so-called ‘anti-woke’ attitudes. When these companies noticed higher levels of engagement in response to extreme material, it seems controls were loosened and algorithms adjusted to increase the amount of such harmful content put in front of people, including young people. Higher engagement levels mean higher advertising revenue and protect a company’s share price. The patchy management of harmful material also seems to have prioritised protecting high profile politicians over vulnerable young people. In other words, hate and division are promoted aggressively in the cause of selling ads online.
Recent research by Labour MP, Liam Byrne identified a ‘media-political complex’ boosting hard right wing political agendas that attracted £170 million in funding over a five year period. Three quarters of that amount went to right wing media platforms: GB News, the Critic and UnHerd. Over £130 million came from just four sources, crypto investor Christopher Harborne (a major donor to the extreme right Reform Party), Dubai-based investment company Legatum, financier Jeremy Hosking, and hedge funder Paul Marshall.
By setting news agendas, the right-leaning political bias in the concentrated private ownership of both legacy and new digital media, tends also then to be magnified by public sector broadcasters. The BBC, for example, has a daily slot on ‘what the papers say’, both online and in its broadcast news. Because there is a predominance of right wing legacy media in the UK that means that the political views of the papers and their billionaire owners are disproportionately amplified. This has editorial knock-on consequences for the views and voices that, once normalised, are in turn further promoted.
Research on the platforming of political parties by UK public service broadcasters, for example, showed a marked bias in favour of the hard-right Reform Party compared to the progressive left Green Party. In 2025 the Greens were referenced just 32 times compared to Reform’s 213. Prominence in news stories was also very unequal. Between January and September that year, Reform led 69 news stories on the BBC and ITV’s News at Ten programmes, and the Green Party only four. On the BBC’s flagship political discussion programme, Question Time, the Greens appeared only about one third as often as Reform.
At some level, however, it seems that the cyber elite are aware of the potential consequences of their influence peddling, and support for politicians who oppose action to keep a safe climate.
The BBC reported that tech billionaires like Mark Zuckerberg are apparently ‘doom prepping’ – building vast, luxury underground bunkers to retreat into. Reid Hoffman, co-founder of LinkedIn, referred to this as ‘apocalypse insurance’.
The acute irony of this response is that, intentionally or otherwise, the alignment of polluter and cyber elites creates ever more febrile conditions for lethal, if not apocalyptic outcomes on a grand scale.
Killing emissions
Dr Stuart Parkinson, SGR’s executive director, speaking at the annual Responsible Science conference, identified the problematic common interests of these two groups, namely: energy intensification, industrial deregulation and militarisation. Also, as seen above, increasingly there are indicators of increasing political alignment.
Global energy supply is still over 80 percent reliant on fossil fuels. Meanwhile, the depletion of reserves together with rising demand is pushing a move to ‘unconventional’ and more polluting fossil fuels like oil shale, gas from fracking, and tar sands. On deregulation, while the fossil fuel industries lobby against climate action, the information and communication technology (ICT) sector pushes back on attempts to regulate the environmental, social and economic threats on everything from data centres to crypto-currencies and social media platforms.
When it comes to militarisation, the fossil fuel and ICT industries are both deeply entwined with the military. Early adopters and heavy users of fossil fuels, many wars are also fought for their control. Not only is the military very dependent on oil, it has multiple exemptions from climate and environmental regulations. At the same time, the application of AI and use of drones and autonomous weapons has rapidly become a feature of conflict. The names of Amazon, Google and Microsoft have all been implicated in the context of the Israeli genocide on Gaza in which the UN Human Rights Council concluded that, “the State of Israel bears responsibility for the failure to prevent genocide, the commission of genocide and the failure to punish genocide against the Palestinians in the Gaza Strip”, says Parkinson. He adds that attempts to update arms control agreements and international humanitarian law to cover the new information technologies have been bogged down in negotiations.
All to sell adverts
Put together there is a toxic cocktail of a far-right political agenda mixing with increased military spending and attempts to deregulate on climate and the environment more broadly, human and workplace rights. A matching accessory to appease these elites - who often espouse libertarian beliefs and the importance of free speech - appears to be the criminalisation of protest that questions or resists their agenda. Part of the deregulation also sees many of these same private sector actors taking on government roles, and links being built with the nuclear industry.
It seems we really are in an apocalyptic conspiracy of shared interests between polluter and cyber elites, driven at a deep level by their need to generate revenue by putting advertising in front of us, thus maintaining overconsumption. All this for the purpose of what Victor Papanek, the great American designer, once immortality described as, “persuading people to buy things they don’t need, with money they don’t have, in order to impress others who don’t care.”
A bewildering array of interventions to course-correct these trends are urgently needed, from treaties to leave fossil fuels in the ground and policies to prevent kleptocratic large-scale tax avoidance by billionaires. But if a single thing is required to create conditions in which such steps can be properly discussed and advanced, it is the protection and expansions of media platforms and information systems that are not privately owned, are free from commercial interests, and independent of advertising revenue.
Backlash
Although this coalition of elite influence exercising its vested interests can seem overwhelming, it does face vulnerabilities. The costs of oil and gas extraction are growing inexorably at the same time as renewables are becoming cheaper and the green economy is growing, the latter becoming more important as an employer and therefore harder to sideline. Similarly, the impacts of global heating and concern about the climate are growing, as is awareness of the deep inequality at the heart of the problem. Rightly, critical gazes fall more and more on the private jets and super yachts.
There is a growing political and cultural backlash against AI stealing not just the jobs, but the mental health and cognitive skills of a younger generation, not to mention the intellectual property of writers, artists and musicians. There is momentum building to establish an internationally recognised ‘AI free’ label for products and services that are authentically human-made, not mediated by AI. Signs and labels are already being used to identify AI-free books, films and websites include, "Proudly Human", "Human-made", “No A.I".
The energy and water costs of data centres also draw negative attention as does the abusive hostility experienced by many online. A significant threat to the economy looms too in an AI-linked threat of a stock market crash.
At the same time the mood of docile acceptance that the nature and shape of the online world, however harmful, is inevitable and cannot be challenged has shifted. In March 2026 the technology giants Meta and Google were found guilty of ‘intentionally built addictive social media platforms’ that were damaging to young people’s mental health. As banal as it may seem, the driver of building these addictive forms of social media was to capture and hold the attention of people for the purpose of putting revenue raising advertising in front of them. Advertising is found to be the dark force lurking behind the scene, catalyst for the design of the addiction engines, the price of which is seen in both mental health and climate-polluting overconsumption.
The future therefore is not smooth for cyber and polluter elites. Those arguing to break their stranglehold on politics have a popular agenda to invoke: jobs, clean air, lower bills, more comfortable homes, technologies better controlled to benefit quality of life and mental health, plus a stronger foundation for international peace.
Standing in the way of this better world is the wealth, exorbitant privilege, media and political control of a very small minority. But the game and its internal connections are increasingly ‘seen’. The irony is that if this wealthy, influential few continue to create their elite hellscape, they will find themselves living in its rubble too, with possibly only a stair climbing, robot vacuum cleaner, with its battery on the blink, for company.
Andrew Simms is Assistant Director of SGR, Co-director of the New Weather Institute, Co-founder of the Badvertising campaign, and Coordinator of the Rapid Transition Alliance.
This article has also been serialised in The Ecologist.
Image credit: Luxury yacht and plane, by Diego F Parra via Pexels.
